AI Avatar Videos

When this makes sense

Not every video needs you on camera.

Weekly "what shipped" recaps, product-release announcements, quick explainers — there’s a class of video where the message is the content and the face is just a delivery mechanism. AI Avatar replaces the recording session without losing the talking-head format.

How it works

Script goes in, MP4 comes out.

From any content item, open Tools → AI Avatar. The script defaults to the first 1500 characters of your transcript or a draft. Pick a D-ID stock presenter (or upload your own licensed avatar image later), hit Generate. D-ID synthesizes audio, lip-syncs it to the presenter, and returns a 1080p MP4 in about 2 minutes. The video drops into your content library ready for the Viral Moments pipeline if you want clips out of it.

• 02Stock D-ID presenters — diverse set, no additional licensing needed
• 02BYOK D-ID key — pay them directly, no Clipflow markup
• 02Works off any content item that has a transcript or draft
• 02Output flows into the standard render + subtitle pipeline

Why Studio only

D-ID calls are expensive relative to other AI features.

Even on BYOK, a 2-minute AI Avatar render costs ~$0.50–1.00 on D-ID’s API. Gated to Studio so the infrastructure cost on our side (webhook handling, storage, retry queue) is covered by the plan price. If you only need this occasionally, a month of Studio is cheaper than a single on-camera shoot.